Expansion of outer cortical CUX2 neurons requires adaptations for DNA repair
Article metadata
Article Date: 01 April 2026
Article URL: https://www.nature.com/articles/s41586-026-10290-4
Article Title: Expansion of outer cortical CUX2 neurons requires adaptations for DNA repair
Article Image: (none provided)
Summary
This Nature study shows that the transcription factor ATF4 is essential for DNA double-strand break (DSB) repair in a subset of embryonic neural progenitors that give rise to CUX2+ upper-layer (L2/3) cortical projection neurons. Conditional deletion of Atf4 in early cortical progenitors (Emx1-Cre;Atf4 fl/fl) causes massive DNA damage, defective ATM phosphorylation, p53-dependent apoptosis of radial glia/early progenitors and selective thinning or loss of CUX2+ L2/3 excitatory neurons. Single-nucleus RNA-seq and molecular assays reveal broad dysregulation of DNA damage response (DDR) pathways in affected radial glial clusters. ATF4 directly regulates key DDR genes (notably Cirbp, Uba52 and Ebf1); Cirbp is required for ATM Ser1981 phosphorylation and proper initiation of DSB repair. Deleting p53 rescues layer thinning by blocking apoptosis, but DNA damage persists. The work links developmental replication/oxidative stress, ATF4-driven transcriptional control of DDR components and the selective vulnerability of expanded supragranular neurons implicated in human neurodevelopmental and neurodegenerative disease.
Key Points
- ATF4 is required in embryonic cortical progenitors for initiating effective double-strand DNA repair and preventing excessive DNA damage.
- Emx1-Cre;Atf4 fl/fl mice display a selective loss and thinning of upper-layer (L2/3) CUX2+ excitatory neurons while deeper layers and interneurons are spared.
- Loss of ATF4 leads to massive accumulation of DNA damage (γH2AX), reduced ATM Ser1981 phosphorylation, and p53-dependent apoptosis in radial glia and early progenitors during E11.5–E13.5.
- Single-nucleus RNA-seq identifies global alterations in DDR-associated gene expression in radial-glial clusters lacking ATF4, without clear activation of canonical UPR/ISR programmes at E11.5.
- ATF4 directly binds and transcriptionally activates Cirbp, Uba52 and Ebf1; knockdown of these genes phenocopies Atf4 loss (increased DNA damage and cell death).
- CIRBP is shown in vivo to be necessary for ATM Ser1981 phosphorylation; Cirbp knockdown reduces p-ATM in radial glia and causes subsequent L2/3 thinning and CUX2+ neuron loss.
- Deleting p53 rescues cortical-layer thinning and survival of L2/3 neurons but does not remove the increased DNA-damage burden, indicating apoptosis is p53-mediated while repair failure remains.
- Findings tie developmental ROS/replication-associated stress and DNA-repair adaptations to the evolution and vulnerability of expanded supragranular (L2/3) neurons implicated in disorders (MS, Alzheimer’s, epilepsy, autism).
Content summary
During corticogenesis, rapid progenitor proliferation produces reactive oxygen species and replication stress that create DSBs. The authors deleted Atf4 across cortical progenitors using Emx1-Cre and found a dramatic, selective depletion of CUX2+ L2/3 excitatory neurons by postnatal stages. The earliest defects appear by E11.5–E13.5 and include a ~30-fold increase in γH2AX, pan-nuclear γH2AX and pKAP1 staining, elevated comet-assay tail DNA and increased cleaved caspase-3. Crucially, ATF4-deficient radial glia show reduced phosphorylation of ATM at Ser1981, indicating failure at the initiation of canonical DSB repair.
Single-nucleus RNA-seq at E11.5 reveals broad dysregulation of DDR-related genes in radial glial clusters; however, gene-set scores for the unfolded protein response (UPR), integrated stress response (ISR), NRF2 targets and global translation were not substantially altered, suggesting a DDR-specific role for ATF4 in this context. Chromatin immunoprecipitation and luciferase assays demonstrate direct ATF4 binding and activation of Cirbp, Uba52 and Ebf1 promoters. Functional studies (lentiviral overexpression/knockdown, aphidicolin-induced DSBs, in utero electroporation) show that these targets are necessary for DNA repair and survival; CIRBP in particular is required for ATM phosphorylation and normal L2/3 formation.
Genetic ablation of p53 in Atf4 cKO mice prevents apoptosis and restores cortical layer thickness, confirming that cell death is p53-dependent though the underlying DNA damage remains high. The authors suggest that recruitment of ATF4-regulated DDR components was an important adaptation enabling expansion of CUX2+ supragranular neurons in mammalian evolution and that the same vulnerabilities underlie selective neuronal loss in neuroinflammatory and neurodegenerative conditions.
Context and relevance
Why this matters: L2/3 CUX2+ pyramidal neurons are disproportionately expanded in evolution and underpin cortico-cortical connectivity tied to higher cognition. This paper provides a mechanistic link between developmental replication/oxidative stress and a bespoke transcriptional DNA-repair programme (ATF4 → CIRBP/UBA52/EBF1) that allows those progenitors to survive and populate upper layers. The results clarify why L2/3 neurons are often selectively vulnerable in diseases (multiple sclerosis, Alzheimer’s disease, TBI-related injury, autism) and suggest DDR components as candidate mediators of neuronal resilience or vulnerability. For researchers, it highlights ATF4 and CIRBP as potential targets to modulate repair capacity in development and disease; for evolutionary biologists, it offers a plausible molecular adaptation tied to cortical expansion.
Why should I read this?
Short version: if you care about why the brain’s upper layers are both huge in humans and oddly fragile in disease, this paper explains the molecular fix — ATF4 — that helps progenitors cope with DNA damage. It’s neat, it’s specific, and it links development, DNA repair and disease in one package. Saves you reading a stack of technical papers: ATF4 → CIRBP (helps ATM) → L2/3 survival. Big implications for developmental neuroscience and neurodegeneration.
Author style
Punchy summary: The authors combine genetics, single-nucleus transcriptomics, molecular assays and in utero manipulations to show that ATF4 is not just a generic stress regulator but a direct transcriptional driver of key DDR elements essential for building the expanded supragranular cortex. This is high-impact for anyone studying cortical development, DNA repair in the nervous system or cell-type selective vulnerability in neurological disease — read the detail if you work in those areas.